Inorg. Chem. 1985, 24, 4158-4164
4158
plexes cis-PtCl2(PNH(CPh0), (4) and c ~ ~ - R C ~ , ( P ( C O ) N H P ~ ) plexes ~ are collected in Table V. (7)(eq 6), and indeed this method of synthesis is the preferred Like the trans analogues 1 and 2, the cis complexes 4 and 7 react with base to give the N-bonded amido complexes. In C/s-Pt( PN(CPh0))z c1S-P tCIz(PN H (CPh 0)) z contrast, however, reaction of 4 or 7 with Dabco in acetonitrile 3 4 solvent rapidly gives the amido complexes at ambient temperature. The respective products 3 and 6 are formed with retention of 16) stereochemistry. This relative ease of intramolecular chloride ion substitution by the amido anion in the cis isomers over the trans c/S-Pt(P(CO)NPh)2 c/s-PtCIz(P(CO)NHPh)2 ones is a consequence of the higher trans influence of the phosphine 6 7 ligand. Registry No. 1, 91410-06-9;2, 98838-49-4;3, 91410-08-1;4, one for complex 7. Complexes 3, 5, 6,10, and 11 do not undergo 91464-49-2; 5,98838-50-7; 6,98919-88-1; 7,98919-89-2; 8,91410-15-0; substitution reactions by added chloride ion; therefore, the initial P(O)NHBz, 98821-87-5; 9,98838-51-8; 10,98838-52-9; 11,98919-90-5; step in their reactions with HCI must involve protonation at PNHBz, 91410-00-3; PCNHPh, 91410-01-4;PNH(CPhO), 91409-99-3; nitrogen. For the platinum amido complexes the stereochemistry P(CO)NHPh, 9 14 10-02-5; P(O)NH-C-C5, 98821-88-6; P(O)NH-C-C,, is retained, but the reaction of either the cis or the trans isomer 98821-89-7; o-Ph2PC6H4CH=NPh9 98821-90-0; o-P~~P(O)C~H~NH~, of Pd(P(CO)NPh), with HCl gives only a single product, 23081-74-5; PhCHO, 100-52-7; Ph2SiH2,775-12-2; o-Ph2PC6H,CH0, PdCl,(P(CO)NHPh), 9 ( b P 19.2). The spectral properties of 4 50777-76-9; PhNH2,62-53-3; o - P ~ ~ P C ~ H ~65423-44-1; N H ~ , PhC(O)CI, and 7 are respectively b p 2.3 (IJ(PtP) = 3646 Hz) and bP 7.6 98-88-4; o-Ph2PC6H4C02H, 17261-28-8; K2PtC14,10025-99-7; Na2PdC14, 13820-53-6. (]J(PtP) = 3737 Hz). N M R and IR spectral data for the com-
K
Contribution from the Chemistry Departments, Ben-Gurion University of the Negev, Beer-Sheva, Israel, and Nuclear Research Center Negev, Beer-Sheva, Israel
Mechanism of Hydrolysis of the Metal-Carbon Bond in a-Hydroxyalkyl-Chromium( 111) Complexes. Effect of Nonparticipating Ligands AMIRA ROTMAN,Ia HAIM COHEN,*Ib and DAN MEYERSTEIN*la Received November 27, 1984 The specific rates of reaction of the free radicals .CH20H, .CH(CH,)OH, and .C(CH,),OH with [Cr"(edta)12-, [Cr"(nta)]-, and trans-[(1,4,8,12-tetraaza~yclopentadecane)(H~0)~Cr~~]~* are reported and compared with the analogous reactions with Cr(H20)2'. The spectra of the a-hydroxyalkybchromium(II1) complexes thus formed are reported and discussed. The rates of hydrolysis of the latter complexes are reported as well as that of cis-[(nta)(H20)Cr"'-CH3]-. The results point out that the electrophile in these hydrolysis reactions is a solvent water molecule and not a cis aqua ligand as earlier suggested. The effects of the pH of the solution and the addition of acetate on the hydrolysis reactions are reported and discussed. The relative stability of alkyl-chromium(II1) complexes in aqueous solutions allows the detailed study of their decomposition mechanisms.2 The large range of stabilities of these complexes and variety of mechanisms of decomposition resulted in extensive studies2 aimed at elucidating the factors affecting the mechanisms and rates of these reactions, which are model reactions to the decomposition of other complexes with u metal-carbon bonds in protic media. Pentaaqua(a-hydroxyalkyl)chromium(III) complexes were shown to decompose in acidic solutions via two pathways. The first is a homolytic decomp~sition:~
-
- d [ [( H 2 0 )SCr'11-CR,R20H]2'] (kH20
dt -k kH30+[H30+l)[ [(H2°)2Cr111-CR,R20H1
"1
(3)
When the measurements were carried out in D 2 0 instead of H 2 0 it was found that both kHIOand d kH30+have large kinetic H / D isotope effect^.^ For the [ H,O']-dependent term the following transition state is commonly accepted
H2O
[(H20),Cr11'-CR,CR20H]2+
+
1
Cr(H20)62+ C R l R 2 0 H (1)
This transition state is in accord with the large kinetic isotope effect and with the observation that bulky R , and R2 substituents deThis mechanism is important only in the presence of scavengers crease kH30+.3 of Cr(H20)62' and/or C R l R 2 0 H ,as K , >> 1. , different transition The second pathway is a heterocyclic hydrolysis r e a ~ t i o n : ~ . ~ For the [acid]-independent term, k ~ 2 0 two states were proposed. H20 1. It was suggested2 that the attacking water molecule is the [(H20),Cr"'-CRlR20H]2+ CrI1laq+ H C R , R 2 0 H (2) ligand bound to the chromium cis to the a-hydroxyalkyl ligand: This reaction obeys in acidic solutions the rate law
-
(1) (a) Ben-Gurion University of the Negev. (b) Nuclear Research Centre
Negev. (2) Espenson, J. H. "Advances in Inorganic and Bioinorganic Mechanisms"; G. Sykes, Ed.; Academic Press: London, 1982;Vol. 1, p 1. (3) Kirker, G . W.; Bakac, A.; Espenson, J. H. J . Am. Chem. SOC.1982,104, 1249. (4) (a) Schmidt, W.; Swinehart, J. H.; Taube, H. J . Am. Chem. SOC.1971, 93, 1 117. (b) Cohen, H.; Meyerstein, D. Inorg. Chem. 1974, 13, 2434.
0020-1669/85/1324-4158$01.50/0
(5) (a) Gold, V.; Wood, D. L. J . Chem. Soc., Dalton Trans. 1981, 2452. (b) Ryan, D. A.; Espenson, J. H. Inorg. Chem. 1981, 20, 4401.
0 1985 American Chemical Society
Inorganic Chemistry, Vof. 24, No. 24, 1985 4159
a-Hydroxyalkyl-Cr(II1) Complexes The arguments in favor of this transition state are as follows:2 a. The water molecule bound as a ligand is a stronger acid than the solvent water molecules. b. No steric effect of bulky R1 and R2 substituents is observed? c. kHZO for the hydrolysis of [(edta)Cr11'XR1R20H]2-is similar to that of [(H20)5Cr111-CR,R20H]2+ though kHIO+for the former is ca.lo5 faster than for the latter.2,6 The effect of edta as a ligand on kHjO+is attributed to the considerably larger stability of Cr"'(edta) than that of Cr3+,.2 The absence of an analogous effect on kH, was attributed to the lack of a cis water ligand in the edta complex, which compensates the expected effect.2 2. On the other hand, it was suggested7 that the transition state for the [acid]-independent term is analogous to that of the [acid]-dependent term, Le.
- -
I
(HZO) S C r [ I I I ) ---C---K ni-'"2I '.n L HA' - ' H
Chart I. Structures of the Complexes Studied
( H p ) sCr :CR1R20H2* c i s + t r a n s water
&-NTACr(H20) :CR1R20H-
cis water
.+
I
H
C-6, 21
0
0
111
The arguments in favor of this transition state are as follows: a. It was found that the rate of hydrolysis of [(H20)5Cr111CR,R20HI2+increases with increasing pH above pH 2.0.' This effect was attributed to the formation of truns-[(H,O),(OH)0 Cr111-CRlR20H]+ and [(H20)3(OH)2Cr111-CR1R20H].It is [EDTA) Cr(H20) :CR1 R20HZt r a n s [15] aneNqCr[H20) :CR1R20HZ+ clear that the remaining cis aqua ligands in the latter complexes are less acidic than those in [(H20)5Cr111-CRlR20H]2f. It is no water coordination t r a n s water therefore more plausible that the attacking water molecule is one from the solvent. the former complex and trans to it in the lattter (nta p nitrilob. The term kHIO for tr~ns-[(H~0)~(CH~C0~-)Crtriacetate and [ 15]ane-N4 1,4,8,12-tetraazacyclopentadecane). C R I R 2 0 H ] + is considerably larger than for [(H20)5Cr111The results clearly point out that a cis aqua ligand has no major CR1R2R3]2+.7This effect is, as that described for the pH effect role in the mechanism of hydrolysis of a-hydroxyalkyl-chromi(see previous paragraph), easier to explain by assuming that the um(II1) complexes. attacking water molecule is one from the solvent. Experimental Section c. The ratio of kH20/kH30+ is considerably larger for the hydrolysis of the [(H20)5Cr11'-CR,R20H]2+complexes than for the Materials. All solutions were prepared from A R grade chemicals, which were used without further treatment, and from distilled water, hydrolysis of the [(H20)5Cr111CRIR2R3]2+ complexes. This effect which was further purified by passing through a milli Q Millipore setup was attributed" to stabilization of the transition state by hydrogen so that its final resistivity was >10 MQ/cm. bonding in the former case, Le.
L
d' 'd" IV
Such an effect is not expected if the attacking water molecule is one of the cis ligands. d. According to this mechanism the increase in kHIO+ for the [ (edta)Cr1I1-CR1R20Hl2-complexes is due to an intramolecular acid attack,6 i.e. [(edta)Cr111-CR1R20H]2-+ H 3 0 + e [(Hedta)Cr111-CR,R20H]- (4) where the acidic form decomposes considerably faster than the alkaline form due to a transition state of the form
V
In order to analyze the plausible role of the cis water ligand we decided to study the mechanism of hydrolysis of the complexes ~is-[(nta)(H~0)Cr~~~-CR,R~0H]and trans-[([ 15laneN4)(H20)Cr111-CR1R20H]2+ as both complexes have one water molecule as a ligand, but due to the structure of the chelating ligand this water molecule is cis to the a-hydroxyalkyl ligand in (6) Ogino, H.; Shimura, M.; Tanaka, N. Inorg. Chem. 1982, 21, 126. (7) Cohen, H.; Meyerstein, D. Inorg. Chem. 1984, 23, 84.
Cr(H20):+ 0.5 M solutions were prepared by dissolving superpure (BDH) chromium powder in 1.0 M HClO, solution continuously purged with Ar or He. The gases were first purified from traces of dioxygen by being bubbled through a washing bottle containing VSO, in dilute H,SO, over Zn amalgam. Solutions of trans-[(H,O)([15]ane-N4)Cr11]2Cwere prepared by mixing deaerated solutions of Cr(H20)? and the free ligand. The ratio of the free ligand concentration to [Cr(H,O)Z'] was always 31.1. The solutions were kept for 0.5 h before irradiation; thus over 95% of the Cr2' was complexed by the ligand.21
Plane, R. A,; Hunt, J. P. J . Am. Chem. Soc. 1957, 79, 3343. Sorek, Y.; Cohen, H.; Mulac, W. A,; Schmidt, K. H.; Meyerstein, D. Inorg. Chem. 1983, 22, 3040. Bakac, A,; Espenson, J. H. Inorg. Chem. 1983, 22, 779. (a) Anbar, M.; Bambenck, M.; Ross, A. B. Natl. Stand. Ref Data Ser. (US.Natl. Bur. Stand.) 1973, NSRDS-NBS 43. (b) Anbar, M.; Farhataziz, N.; Ross, A. B. Ibid. 1975, NSRDS-NBS 51. (c) Farhataziz, N.; Ross, A. B. Ibid. 1977, NSRDS-NBS 59. Lati, J.; Meyerstein, D. J . Chem. Soc., Dalton Trans. 1978, 1104. Misckei, K.; Debreczeni, F.; Nogypal, I. J . Chem. Soc., Dalton Trans. 1983, 1335. Perin, D. D. "Stability Constants of Metal Ion Complexes, Part B"; Wheaton, and Co.: Oxford, England, 1979; p 449. Matheson, M. S.;Dorfman, L. M. "Pulse Radiolysis"; MIT Press: Cambridge, MA, 1969; Chapter 6. In solutions containing C,H,OH and CH(CH,)20H ca. 13% of the free radicals formed are .CH2CH20Hand .CH2CH(OH)CH3,respectiveI Y . " ~ The corresponding alkyl-chromium complexes decompose via a fast @-eliminationproce~s''~ and thus did not interfere in this study. (a) Asmus, K. D.; Mockel, H.; Henglein, A. J . Phys. Chem. 1973, 77, 1218. (b) Cohen, H.; Meyerstein, D., submitted for publication. See, for example: Buxton, G. V.; Sellers, R. M.; McCracken, D. R.J . Chem. Soc., Faraday Trans. I 1976, 72, 1464. Jubran, N.; Ginzburg, G.: Cohen, H.; Koresh, Y . ;Meyerstein, D. Inorg. Chem. 1985, 24, 251. k o , . + c ~ , c ~has o not been reported; however, koH+CH OH)2 = 2 X lo9 M-' s-I ,I I and koH+CH3CH2CH CHO = 3.5 X lo9 M-' s-I.~( This the rate for ea; + &(edta)-;" the rate for the nta complex might be slightly larger. Samuels, G. J.; Espenson, J. H. Inorg. Chem. 1982, 21, 126.
Rotman et al.
4160 Inorganic Chemistry, Vol. 24, No. 24, 1985 All solutions containing Cr(H20)t' and/or L5Cr"'-CRIR20H complexes were handled by the syringe technique in order to avoid contact with dioxygen. Techniques. Spectrophotometric measurements were carried out with a Spectronic 2000 Bausch and Lomb spectrophotometer. pH determinations of solutions containing Cr(H20)62' were carried out by immersing the combined electrode into the syringe while Ar was bubbled through the solution. The concentration of C r ( H 2 0 ) p solutions was determined by oxidation to Cr(H20)63+ and measurement of the optical density at 408 nm where c(Cr(H20)2+)= 15.6 M-' cm-1.8 In dilute solutions the concentration of chromium was determined with an atomic absorption spectrophotometer. Pulse-radiolysis experiments were carried out at the linear electron accelerator facility of the Hebrew University of Jerusalem; 0.2-1.5-ps, 5-MeV, and 200-mA pulses were used. The dose per pulse was in the range 500-4000 rad/pulse. The experimental setup as well as the techniques used for evaluating the results has been described elsewhere in detail.4.9 Preparation of the a-HydroxyakylChromium(II1)Complexes. Two general methods were used to prepare the a-hydroxyalkyl-chromium(1II) complexes. The key reaction in both methods is4 L5Cr"H20
+ .CR,R20H
-+
[L5Cr"'-CRIR20H] + H 2 0
(5)
Thus both methods are aimed at producing .CRIR20H free radicals in the presence of L5Cr"H,O. The difference between the two techniques is mainly their time resolution. The Modified Fenton Reagent. This technique was used for the study of the kinetics of decomposition of a-hydroxyalkyl-chromium complexes with a half-life longer than 2 min, i.e. of the [(nta)(H,O)Cr"'CR,R,OH}- complexes. A deaerated solution containing Cr(H20)2', nta, CH3C02Na,and H C R , R 2 0 H at the required pH is mixed in a syringe with a deaerated solution containing H20,. Upon mixing the following reactions occur:
H P
[Cr"(nta)(H20)2]-+ H 2 0 2 orlo
[Cr"'(nta)(H,O),]
-
Cr(H20)2+ + H 2 0 2
-
*OH + HCRIR2OH
k7 > 8 [Cr"(nta)(H20),]-
X
H20
Crlllaq
+ .OH
-
C r ( H 2 0 ) 2 ++ .CR,R,OH
+ CH3C02-
H2O
+ CRIR2OH
k , , = 7.5
-+
X
H20
[(nta)(H20)Cr"'-CRlR20H](8)
+ .CH,CO,-
s-I
p
IO6 M-'
X
(11)
= 1 M was reported:
+ nta3- 2 [Cr(nta)(H20),]-
K = 3.2
(12)
l3
At a lower ionic strength, as in this study, a considerably higher stability constant is expected. The pK,'s for nta are pK,, = 9.5, pKa2= 7.2, and pKa3 = 1.7.14 From these values the apparent stability constants of [Cr(nta)(H,O),]- at pH 4.0 and 5.0 are 10 and 100 M-I, respectively. In order to avoid a competition of reaction 9 with reaction 8, very high concentrations of nta are required in this pH range. These concentrations are impractical due to the requirement that [HCR,R,OH] > 30[nta]. Thus under our experimental conditions reaction 9 competes with reaction 8. The [(H20)5Cr111-CR,R20H]2+ complex thus formed decomposes via two competing pathways:
-
[(H20)5Cr"'-CRlR20H]2+
k>3
X
H P
Cr'",,
+ HCRlR20H
IO-* s-, at pH >4.07
[(H20)5Cr"'-CRlR20H]2++ nta3-
H,O
Y , e-
e a i , .H, .OH, H,, H 2 0 z ,H 3 0 +
(15)
The yields of these products are Gea- = 2.65, GOH= 2.65, GH = 0.60, GHi = 0.45 and GH2O2= 0.75. (G is Jefined as the number of molecules of each product per 100 eV absorbed in the s o l u t i ~ n . ' ~In) concentrated solutions the yield of the primary free radicals is often somewhat higher and that of H2 and H 2 0 2somewhat 10wer.l~ In neutral N,O-saturated solutions, [ N 2 0 ] = 2.2 X M, the following reaction occurs: N 2 0+ e a i
HP
N2
+ .OH
k = 8.7 x lo9 M-I s-l
(16)
thus all the primary free radicals aretransformed into .OH and .H free radicals. The -OH radicals react in the systems studied via reactions 7, 10, and 11 and the .H atoms via
+ HCRIR20H
-
+
H2
+ *CR,R*OH
(17)
k > 1.6 X IO6 M-' s-'
+ CH,CO< H 2 + C H 2 C 0 2 - k = 3.5 X lo5 M-l s - I I 1 *H + N(CH,COz-)3 H, + N(CH2CO2-),(*CHC02-)
-
k = 7.5
X
IO6 M-' s-I
(18) (19)
Thus in N20-saturated solutions, under our experimental conditions, all the free radicals are transformed into the desired .CRlR20H free radicals.I6 We observed that when solutions containing [Cr(nta)(H,O),]- are saturated with N,O their color changes from blue to violet; Le., the complex is oxidized to Cr(nta)(H,O), probably via [Cr"(nta)(H,O)J
+ N,O
H20
[(H20)(nta)Cr'v=O]-
followed by [(H,O)(nta)Crlv=O]-
+ [Cr"(nta)(H20),]' + 2H30+
+ N,
(20)
-
2 [Cr"'(nta) (H20),] (2 1)
at pH 4.012
The stability constant of [Cr(nta)(H,O),-] at Cr(H,O)P
(10)
+ N(CH2CO,-),(.CHCO,-)
lo8 M-l
From these data it is clear that reaction 2 did not interfere with our measurement of the specific rates of reaction 14. As in many experiments, see below, the nta concentration was relatively low, and it is clear that k,,[nta3-] > k2. in analogy to similar reports for the edta system6 Pulse-RadiolysisExperiments. This technique was used for measuring the specific rates of reaction 8 and the rates of decomposition of [L5Cr"'-CRlR20H] complexes with short lifetimes. This technique was also used to determine the molar absorption coefficients of the [L5Cr'1'-CR,R20H] complexes and for measuring the absorption spectrum of [([ 15lane N4)(H20)Cr"'-C(CH3)20H]2+. The radiolysis of water and dilute aqueous solutions can be summed up byI5
(7)
k , , = 8.5 x 107 M-I s-111 *OH + N(CH,CO
